With the rapid development of my country's railway industry, no defects on the track surface are a prerequisite to ensure the normal operation of the train. Using deep learning instead of humans to detect defects on track surfaces can greatly improve the efficiency of detection while ensuring the safety of workers. However, in real life, railway workers will quickly repair and replace once the rail fails. Although this measure ensures the smooth operation of the train, it will bring difficulties to the collection of datasets, resulting in the lack of data set images of tracks with surface defects. In this paper, Gan is used to extend the dataset and then EfficientDet algorithm is used to detect and recognize the defects on the track surface. By comparing the efficiency and accuracy of the efficientD0, D3 and D6 algorithms, the final one was EfficientDet-D3. the experiment showed that the D3 version could achieve faster detection speed while ensuring the accuracy.
Track Surface Defect Detection Based on EfficientDet
Lect. Notes Electrical Eng.
International Conference on Electrical and Information Technologies for Rail Transportation ; 2021 October 21, 2021 - October 23, 2021
Proceedings of the 5th International Conference on Electrical Engineering and Information Technologies for Rail Transportation (EITRT) 2021 ; Kapitel : 7 ; 56-66
2022-02-19
11 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Track Surface Defect Detection Based on EfficientDet
British Library Conference Proceedings | 2022
|Object Detection Using Vision Transformed EfficientDet
IEEE | 2023
|Object Detection in Autonomous Maritime Vehicles: Comparison Between YOLO V8 and EfficientDet
Springer Verlag | 2023
|Track Surface Defect Detection Based on Image Processing
British Library Conference Proceedings | 2018
|